Yeah, i've thought of that before i've setup the backup. I currently have 10 backup jobs, like this:
# grep Concurrent * bacula-dir.conf: Maximum Concurrent Jobs = 10 bacula-fd.conf: Maximum Concurrent Jobs = 10 bacula-sd.conf: Maximum Concurrent Jobs = 10 So this shouldn't be be an issue. If I do the status of the Direct, I get the following: Running Jobs: Console connected at 02-Jul-05 19:18 JobId Level Name Status ====================================================================== 179 Full BackupCatalog.2005-07-03_03.20.00 is waiting execution 178 Full kaufman-Full.2005-07-03_02.30.02 is waiting execution 177 Full dyno-Full.2005-07-03_02.30.01 is waiting execution 176 Full whale-Full.2005-07-03_02.30.00 is waiting execution 175 Full heart-a-Full.2005-07-03_00.05.01 is waiting execution 174 Full filo-Full.2005-07-03_00.05.00 is waiting for higher priority jobs to finish 173 Full remote.2005-07-02_19.19.16 is running ==== As you can see, the full backup job 174 is waiting for higher priority jobs to finish. However, job 174 (filo) has the highest priority out of all and remote (173) has the lowest priority. In anyways, assuming that simultanious jobs are allowed, this should not have an effect, as those two jobs have different volumes and run on different drives and tapes. So there seems to be some bugs with executing more than one job at a time. Has anyone else experienced similar issues? -- Andrei On Sun, 2005-07-03 at 22:50 +0200, Michel Meyers wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Andrei Mikhailovsky wrote: > > However, the local backup jobs are still wait for the remote one to > > finish. So, I will have about 3 days of missed backup, which I can't > > really afford. > > > > Does anyone know what I am doing wrong? > > Read up on 'Concurrent' jobs in the manual > http://www.bacula.org/rel-manual/Tips_Suggestions.html#SECTION0003118000000000000000 > and search the mailing list archives for it if you want. > > The trick is to make sure that there is no bottleneck in your > configuration (ie, you specify 10 concurrent jobs in the SD but only one > in the directory and bang, you've reduced the maximum of simultaneous > jobs to 1) and you have to double and triple check that it it set up > properly. > > Greetings, > Michel > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.1 (MingW32) - GPGrelay v0.959 > > iD8DBQFCyE+D2Vs+MkscAyURAmvZAKDYp0QBeQFpkkrjC0juyEksc+XWqwCfZhU4 > KQ7ssXHWi6SOdgN+fRv62Ec= > =8mb7 > -----END PGP SIGNATURE-----
signature.asc
Description: This is a digitally signed message part
